Professional Documents
Culture Documents
tmp3539 TMP
tmp3539 TMP
tmp3539 TMP
Dipartimento di Scienze Fisiche, Matematiche e Informatiche, Via Campi 213/A, 41125 Modena, Italy
CNR Istituto Nanoscienze, S3, Via Campi 213/A, 41125 Modena, Italy
Dipartimento di Scienze della Vita, Universit di Modena e Reggio Emilia, Via Campi 287, Modena 287, Italy
d
CNR Istituto di Biosica, Via De Marini 6, 16149 Genova, Italy
b
c
a r t i c l e
i n f o
Article history:
Received 30 August 2014
Received in revised form 18 December 2014
Accepted 2 January 2015
Available online xxxx
Keywords:
Atomic Force Microscopy
Neurosteroid
Critical uctuation
Allopregnanolone
Liquid ordered domain
a b s t r a c t
Amphiphilic molecules which have a biological effect on specic membrane proteins, could also affect lipid bilayer properties possibly resulting in a modulation of the overall membrane behavior. In light of this consideration, it
is important to study the possible effects of amphiphilic molecule of pharmacological interest on model systems
which recapitulate some of the main properties of the biological plasma membranes. In this work we studied the
effect of a neurosteroid, Allopregnanolone (3,5-tetrahydroprogesterone or Allo), on a model bilayer composed by the ternary lipid mixture DOPC/bSM/chol. We chose ternary mixtures which present, at room temperature, a phase coexistence of liquid ordered (Lo) and liquid disordered (Ld) domains and which reside near to a
critical point. We found that Allo, which is able to strongly partition in the lipid bilayer, induces a marked increase
in the bilayer area and modies the relative proportion of the two phases favoring the Ld phase. We also found
that the neurosteroid shifts the miscibility temperature to higher values in a way similarly to what happens
when the cholesterol concentration is decreased. Interestingly, an isoform of Allo, isoAllopregnanolone (3,5tetrahydroprogesterone or isoAllo), known to inhibit the effects of Allo on GABAA receptors, has an opposite
effect on the bilayer properties.
2015 Elsevier B.V. All rights reserved.
1. Introduction
Many commercially available drugs are directed to membrane proteins. Membrane proteins are strongly coupled to the lipid component
of the biological membrane both by specic chemical interactions [1]
and by longer range, aspecic physical interactions [2]. The most commonly accepted mechanism of drug action relies on specic and saturable interactions between the exogenous molecules and membrane
proteins. This mechanism has been demonstrated in many cases and
the binding sites for many ligands and their receptors have been clearly
identied [3,4]. A well accepted evidence for the specicity of the interaction between drugs and proteins lies in the different pharmacological
properties of the enantiomeric forms of a drug [5]. These differences are
typically related to the chiral structure of proteins and to the specic
docking of the drugs to the proteins. Nevertheless, the effect of a drug
on a membrane protein cannot be studied independently of the presence of the lipid bilayer whereas it is possible to study the effect of
drugs on pure lipid systems. The kinetics of the drug/membrane protein
http://dx.doi.org/10.1016/j.bbamem.2015.01.002
0005-2736/ 2015 Elsevier B.V. All rights reserved.
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
Besides these aspects, the up to date view of the biological membrane behavior involves lateral heterogeneity in the organization of
lipids and proteins. Specic domains with a lateral extension in the
order of tens of nanometers could be present in the membrane [9].
Many proteins could preferentially partition in these domains and the
partitioning could affect their interactions and their activation or deactivation of signaling pathways [10,11]. The membrane organization in
nanometer-sized lipid aggregates could be related to its thermodynamic conditions [12] considering also the non-equilibrium situation of a
biological cell membrane. The interaction of drugs with the membrane
could in turn affect its thermodynamics and alter the bilayer organization [13]. These changes could have strong effects on the activity of
membrane proteins which preferentially associate with specic
domains. Accordingly, it would be very important to study how
the insertion of exogenous molecules in a lipid bilayer affects its
thermodynamics.
Dealing with the possible role of drugs in affecting the functional
activities of biological membranes by an aspecic mechanism, one of
the most studied effects is the interplay between anesthetics and lipid
bilayers [1418]. The longstanding debate between the specic effects
of anesthetics on membrane proteins and their indirect action, mediated by changes they may produce in the lipid bilayer properties, has its
roots in the MeyerOverton rule. This rule states that the activity of anesthetics is strongly related to their partition coefcient in lipid bilayers.
Among the mechanisms related to aspecic interactions, the lateral
pressure prole change across the lipid bilayer has been proposed as a
way to affect efciently the equilibrium distribution of the membrane
proteins in their different conformational states and the rate constants
for their transitions [19]. Moreover, recent studies on model bilayers
showed that the interaction of anesthetics with lipid bilayers affect
their thermodynamic state by changing the relative distribution between different phases [2022] or by changing the temperature of the
bilayer at which a separation in two liquid phases occurs [13]. Another
interesting class of molecules whose interactions with lipid bilayers
could be relevant for their activity is exemplied by neurosteroids
[NSs] [23] which are endogenous molecules able to modulate the activity of ion channels and relevant in the propagation of electrical signals in
the nervous system [24]. Their action is typically explained by an allosteric interaction with a membrane protein, i.e., the GABAA receptor Cl
channel, that, affecting the receptor conformational states, produces
variation in the channel open time that ultimately results in a changed
activity [25,26]. Their activity is usually detected at very low concentrations (in the nM range) and this aspect initially suggested a highly specic interaction with the corresponding membrane proteins [27].
Evidences for a specic NS binding site on GABAA receptor were provided by several groups. Using site-directed mutagenesis, single residues in
the receptor protein that inuence NS regulation of GABA receptor have
been identied [2830]. However, NSs are strongly lipophilic and, according to their structure, in some cases they have a partition coefcient
which can produce, from a nM concentration in aqueous solution, a M
concentration inside the lipid bilayer [31]. In this case, a highly specic
docking mechanism would not be required to explain their effect at a
very low solution concentration. Moreover, it has been found that
their docking site to the membrane protein could be located in their
intramembraneous portion [27]. Accordingly, the effects of NSs could
be also related to their partition and diffusion inside the lipid bilayer.
It is also well known that, at high concentrations, NSs which modulate
the activity of GABAA receptors can, independently from the GABA presence, activate the receptors [32]. This direct gating effect of the NS has
typically slow kinetics which has been connected to its accumulation
in the lipid bilayer [33]. The gating kinetics could be attributed both to
the required increase of NS concentration inside the bilayer to bind
the channels and to the increasing modication of the lipid bilayer properties as the NS concentration inside the bilayer increases.
In this work we studied the effect of two NSs, Allo and one of its isoforms, isoAllo, on a ternary lipid model-membrane containing a natural
sphingomyelin extract. We concentrated on the DOPC/bSM/chol mixture with different lipid proportions. bSM is a mixture per se, even if it
is mainly composed (~50%) by a 18:0 fatty acid chain. Accordingly, the
mixture will be considered as a pseudo-ternary lipid mixture. This
lipid combination is considered a very representative case for the studies of phase behavior of ternary lipid bilayers [34]. One of its main characteristics is related to the fact that the typically used 1:1:1 molar
mixture is, at room temperature, very near to a critical point. In fact,
the presence of critical points in the thermodynamics of biological
membranes is nowadays considered one of the possible explanations
for the existence of small and dynamic domains usually called lipid
rafts [12]. Here, to study the effects of Allo and of its isoform isoAllo
we exploited Atomic Force Microscopy (AFM) after forming Supported
Lipid Bilayers (SLBs) of the specic lipid composition. While Allo potentiates GABAA receptor activity in the presence of GABA and, at high concentrations, directly gates the GABAA receptor channel [25,35], isoAllo is
devoid of modulatory activity but acts as a non-competitive antagonist
of Allo [36]. In particular, isoAllo decreases the effect of potentiation
by Allo and the effect is particularly evident at high GABA concentrations. Herein, we focussed on the modication of the thermodynamic
state of the bilayer as the concentration of the NSs in the imaging chamber increased, measuring the relative proportions of different phases
which might be present in the bilayer. We also used a temperaturecontrolled stage to investigate the behavior of the lipid bilayer as a function of temperature.
2. Materials and methods
2.1. Lipid bilayer formation and neurosteroid injection
The lipids 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC),
Sphingomyelin (Brain, Porcine) (bSM) and cholesterol were purchased
from Avanti Polar Lipids (Alabaster, USA) and used without further purication. Allopregnanolone was purchased from Sigma-Aldrich and
isoallopregnanolone was a kind gift from Dr A. Guidotti (refer to [25]
for details on this molecule). Specic lipid mixtures were prepared by
mixing chloroform lipid solutions in the desired molar amount. Chloroform was then evaporated under a ow of nitrogen while being heated
in a water bath at 50 C. Thereafter, the sample was kept under vacuum
(102 mbar) for at least 2 hours in order to remove the remaining chloroform. Then lipids were rehydrated in a buffer solution (150 mM KCl,
8 mM Hepes, pH 7) to obtain a lipid concentration of 0.12/0.06 mg/ml.
The sample was sonicated at room temperature for 15 min resulting
in a homogenous lipid suspension.
SLBs were prepared by the vesicle fusion technique [3739]. Briey,
immediately after sonication of the lipid suspension, 70100 l of the
suspension was deposited onto a freshly cleaved mica sheet (SPI Supplies/Structure Probe, Inc., USA) xed on a PTFE disc attached to a
metal disc. The lipid suspension was incubated for 15 min at a temperature above 40 C and then the sample subjected to extensive rinsing
with the imaging buffer. The sample was then slowly cooled to 25C.
Small amounts of the NSs (106 M concentration in the same buffer
used for AFM imaging, diluted from a 10 2 M DMSO solution) were
injected in the imaging chamber in order to reach the desired nal concentration. After each injection, we waited for about 1015 min in order
to reach an equilibrium condition for the bilayer before acquiring images. It is important to stress that control experiments on lipid bilayer
patches prepared in the same conditions but not exposed to NSs conserved good time stability in their properties. All the reported variations
after the insertion of the exogenous molecules are therefore ascribable
in large part to their effect on the bilayers.
2.2. Atomic Force Microscopy
Atomic Force Microscopy imaging was performed with a Bioscope
I microscope equipped with a Nanoscope IIIA controller (Veeco
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
Fig. 1. Phase diagram for the ternary lipid mixture DOPC/bSM/chol based on Ref. [48]. Two
regions of phase coexistence are highlighted: the Liquid Disordered/Liquid Ordered phase
coexistence region (Ld + Lo) and the triangular three phase coexistence region in which
also a Solid Ordered phase is present (Ld + Lo + So). The phase diagram represents a
slice at about 25 C of the complete phase diagram. The straight dashed lines inside the liquidliquid phase coexistence regions represent schematically the tie-lines.
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
Fig. 2. AFM images of a DOPC/bSM/chol 1:1:1 supported lipid bilayer at 25 C. a) The image
shows homogeneous lipid bilayer patches (*) and smaller lipid bilayers in a phase coexistence state (#). b) Height difference between the homogeneous bilayer and the domains
in a patch with phase coexistence. The height of the homogeneous bilayer is intermediate
between the two domains in the small patch. The line section refers to the white line on
the image.
after Allo perfusion. The immediate effect is represented by a strong increase of the lateral area occupied by the lipid bilayer. When we focus
the attention on small patches such as those shown in the lower panel
(sequence d, e, f) of Fig. 3, we nd that the area increase is typically
1520%. This behavior is consistent with what was found in the case
of GUVs of the same lipid composition exposed to a solution containing
Allo and studied by the Micropipette Aspiration Technique in which Allo
uptake induces a global area increase of the bilayer [57]. Even if the two
model systems, GUVs and SLBs, have different characteristics, in some
cases they can be prepared so to provide similar results dealing with
their phase organization, as shown in a recent report [58]. Moreover it
has been reported that a SLB undergoes the same relative area variation
of an unsupported lipid bilayer when it crosses its main phase transition
temperature [59]. Accordingly, a similar behavior between a SLB and a
GUV provides support to the interpretation of changes observed to a
SLB.
Besides area increase, Allo induces also a reorganization of the lipid
domains. It is important to point out that, in the absence of Allo, under
similar imaging conditions, the domains are rather immobile. The reorganization is characterized by a relative size increase of the regions in
the Ld phase. It is to be stressed that the lipid mixture at issue is very
near to a critical point for this temperature and, accordingly, the relative
distribution between the different phases is near 50% for each phase. An
average estimation for the change in the relative presence of the domains shows that the Ld phase relative increase is (10 4)% after
reaching 300 nM Allo concentration. Table 1 summarizes the results
showing the relative proportions of the different phases and the total
area variation (see Supporting information for further details). At the
level of the area of single domains, the insertion of Allo produces a
small size increase also of the Lo domains. This behavior could be the result of the fact that Allo produces a global change of the bilayer state, as
observed also for the introduction of trace amounts of uorescent lipids
in ternary lipid mixtures [60]. Fig. 3g shows the analysis of the height
difference between the domains when the NS is added to the lipid bilayer. The presence of Allo leads to an increase of the height difference between the domains. This behavior is consistent with an increased
composition difference for the domains and would correspond to a
shift in the phase diagram of Fig. 1 towards a region of decreased cholesterol amount on the basis of the reported positive slope of the tie-lines
in the two-liquid phase coexistence region [48]. Moreover, the height
difference variation results by a variation of both the Lo and the Ld
phases with respect to the mica substrate. In fact, the Lo phase thickness
Fig. 3. Effect of Allo on the lipid bilayer. a) Lipid bilayer patches composed by the lipid mixture DOPC/bSM/chol 1:1:1 in the absence of Allo; b) The same sample area as in a) after the
injection of 50 nM Allo; c) The same sample area after the injection of 100 nM Allo; df) Sequence of images at higher magnication representing the structure of a lipid bilayer patch
presenting a phase coexistence state: d) control (no Allo); e) 50 nM Allo; f) 100 nM Allo. g) Cross sections corresponding to the image sequence df) (white line in d)).
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
Allo
iso-Allo
Lo
Ld
Lo
Ld
Initial
300 nM
(43 5)%
(57 5)%
(50 3)%
(50 3)%
(33 4)%
(67 4)%
(52 3)%
(48 3)%
+(24 5)%
(14 3)%
increases after the introduction of Allo while the Ld phase slightly decreases. This behavior once again supports the idea that Allo induces a
global transformation in the lipid bilayer. Considering these observations, the effect of Allo on this lipid bilayer is equivalent to decreasing
the cholesterol content or reducing the interaction between cholesterol
and bSM. In this context, it is interesting to note that the effect of Allo on
the GABAA receptor has been found related to the amount of cholesterol
in cells [61,62]. In particular, it has been demonstrated that the effect of
steroidal GABA potentiators was reduced by increasing cholesterol
amount in cells. At the same time the amount of cholesterol in cells is
also able to modulate the GABA potency, with a reduction of the agonist
potency after cholesterol depletion [63]. The results we obtained in our
study suggest that NS could compete with cholesterol inside the bilayer
and could alter its effect and its interaction with high melting temperature lipids.
As stated above, the hypothesis of the presence of a critical point in a
biological membrane is nowadays considered very appealing because it
could explain peculiar behaviors of these thermodynamic systems such
as the existence of small and uctuating domains [12]. From a biological
point of view it is probable that a single critical point might represent a
too specic situation to have functional relevance, but it must be
stressed that multidimensional order parameters and several critical
points might be present in these models [64]. Therefore we decided to
Fig. 4. Effect of Allo on the miscibility border of the DOPC/bSM/chol 1:1:1 bilayer. a) Schematic representation of the phase diagram as a function of temperature [based on Ref. 46]. The
continuous lines represent the upper border of the miscibility region for different temperatures, whereas the dashed line shows the movements of the critical points as a result of temperature variation. As the temperature increases, the miscibility border shifts to lower regions of the Gibbs triangle. b) DOPC/bSM/chol 1:1:1 bilayer at 27 C in a homogeneous phase.
ce) the same bilayer as in b) after the addition of c) 250 nM Allo, d) 350 nM Allo and e) 450 nM Allo. f) Height difference between the phases formed after Allo addition as a function
of its concentration.
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
study the behavior of the lipid mixture at issue near to its critical point.
The identication of critical points in similar lipid mixtures has been
performed by uorescence microscopy [42] as well as by AFM [46].
Here we increased the temperature of the imaging chamber until we
completely removed the phase coexistence. We did not measure the parameters which are required to ascertain the real critical position of the
lipid bilayer because we were mainly interested in the miscibility transition temperature. Increasing the temperature, the miscibility border in
the pseudo-ternary phase diagram changes (see the scheme reported in
Fig. 4a which has been reproduced as a sketch from ref [46]). A bilayer,
initially in the phase coexistence region, is driven to a homogeneous
phase by increasing temperature. Approaching the miscibility border
near to a critical point involves the presence of domains with very similar composition and a very low line tension favoring the formation of
small domains. Those domains are nevertheless visible by AFM. Outside
the phase coexistence region and near to a critical point, compositional
uctuations should play a dominant role. Due to the limited time resolution of AFM, these uctuations could result unattainable to the imaging capabilities of AFM or, at least, AFM images cannot be considered as
valuable snapshots of the bilayer state. After getting a homogeneous bilayer, as shown in Fig. 4b, we added Allo to the lipid bilayer. The progressive addition of the NS leads to the reappearance of very small
domains, as shown in Fig. 4c that corresponds to 250 nM Allo concentration. Upon further increase in the concentration of the exogenous molecule, the different domains in the bilayer progressively increase their
height difference, according to an increasing composition difference between the domains (Fig. 4de). A plot of the height difference between
the two different phases as a function of Allo concentration is reported
in Fig. 4f. If, at a constant Allo concentration (in this case 450 nM), we
decrease the temperature, we again observe an increase in the height
difference (Fig. 5ad). This behavior suggests that Allo addition is equivalent to what would happen by decreasing the temperature for the lipid
bilayer (like in the freezing point depression case). It is interesting to note
that the application of the NS to a lipid bilayer initially near to a critical
point maintains a situation compatible with the neighborhood to a similar point, as manifested by the formation of small and uctuating domains. This behavior is also conrmed by the fact that, after the
insertion of Allo, the relative fraction of the two phases is again near
to 50%.
Fig. 5e reports the height difference between domains for a pure
lipid bilayer and a lipid bilayer exposed to 400 nM Allo concentration
as a function of the temperature. For the pure lipid bilayer, the height
difference approaches a value of 0 nm for a temperature of 30 C as expected for a critical point at such temperature. The height difference can
Fig. 5. Effect of the temperature in the presence of 450 nM Allo: a) 27 C; b) 25.8 C; c) 24.5 C; d) 23.3 C. e) Height difference between the Lo and Ld domains in the case of a pure DOPC/
bSM/chol 1:1:1 supported lipid bilayer (lled squares) and in the case of the same bilayer exposed to a 400 nM Allo solution (open squares). The data have been tted with the function
IT TcI. The obtained values for are 0.28 for a pure lipid bilayer and 0.26 for a bilayer in the presence of Allo.
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
putatively solid domains, as it can be seen from Fig. 6c, d which corresponds to 400 nM Allo. Indeed, the Lo domain on the right and lower
portion of Fig. 6c shows the presence of a region with a slightly larger
thickness. The identication of So regions by AFM in ternary lipid mixtures can be elusive due to the very small height difference with respect
to Lo domains. Fig. 6d shows an averaged line section of the domain in
Fig. 6c whereas Fig. 6e shows the image acquired by the phase signal
of tapping-mode AFM on the same region of Fig. 6c. The phase signal
produces a visible contrast when the tip of the AFM is interacting with
regions of different mechanical properties [67,68]. The variation of the
phase contrast on the higher domain in Fig. 6e could correspond to
the presence of a So region that is mechanically more rigid than the Lo
phase.
In the debate on the possible effects of drugs on biological membranes a point that is usually considered crucial to distinguish between
a specic effect on membrane proteins and an aspecic action on the
lipid bilayer physical properties is the difference in stereoisomers' activity [69,70]. More generally, if colligative properties are considered, enantiomers of the same molecule should not produce different effects
on an achiral lipid bilayer. Accordingly, if two enantiomers change differently the activity of a membrane protein, such as in the case of the
two enantiomers of Allo on the GABAA receptor [71], the mechanistic
explanation should foresee a specic interaction of the drug with the
membrane protein. However, it is not trivial to study the complete effect
of a molecule on a lipid bilayer, because it could affect properties which
are not easy to study experimentally, such as the lateral pressure prole
inside the bilayer [16], a property still waiting for an appropriate experimental technique able to provide its measurement. In general terms, it
is important to stress that the lipid bilayer cannot be considered an achiral environment, especially if cholesterol is one of the components [72].
Dealing with the phase organization of lipid bilayers, it is important to
study if isomers of the drugs could differently alter the phase state of
the bilayer inducing the prevalence of one phase over other coexisting
phases.
In this work we considered also the effect of isoAllopregnanolone, a
3 isoform of Allo differing only for the spatial orientation of an OH
group, on a lipid bilayer of the same composition. It is important to
stress that isoAllo is not a mirror image of Allo, and it is reasonable to expect a different interaction with a lipid bilayer [73,74]. However, given
that the functional effect of isoAllo is that of inhibiting the activity of
Allo in a non-competitive way [36,75], it is interesting to study if the
two molecules have also different and somehow opposite effects on
the lipid bilayer. Fig. 7 shows the effects of the perfusion of isoAllo on
a DOPC/bSM/chol 1:1:1 supported bilayer. It is evident that in this
case the insertion of this NS in the bilayer induces a decrease of the overall lipid bilayer area. Moreover, an analysis of the relative proportion of
the two phases shows that the addition of isoAllo induces an increase of
the Lo fraction over the Ld one. The variation of the Lo fraction for a
300 nM isoAllo concentration is (2 1)%. Table 1 reports a comparison
between the effects of the same concentration of the two molecules on
the Lo and Ld and average area variation of the here considered ternary
lipid mixture. Interestingly, the interaction of GABAergic phenols on
lipid bilayers has been recently studied by 1H NMR [76] and it has
been found that these molecules strongly interact with the lipid bilayer
partitioning near to the polar group and the rst part of the acyl chains.
At the same time, these molecules favor a tighter packing of the lipid bilayer. It could be that the insertion of isoAllo in the DOPC/bSM/chol mixture induces a decreased repulsive force between the lipids involved in
the Lo phase and a decrease of the overall membrane area.
Considering the results obtained in the present study, we could
speculate that the antagonistic effect of isoAllo on Allo could also be related to an opposite effect on the lipid bilayer.
We also investigated the effect of isoAllo when the lipid bilayer is
just outside the phase coexistence region. In this case we found that,
up to 400 nM isoAllo concentration, no effect was evident from the
AFM images. This is at variance with what happens in the case of Allo
which stabilizes the phase separation of the bilayer. In the case of isoAllo
it could be that the addition of the exogenous molecule shifts the system
to the homogeneous phase.
In order to get an idea of the effect of the NSs on the global acyl chain
order in the bilayer we followed by transmission FT-IR spectroscopy the
position of the absorption maximum due to the symmetric \CH2
stretching mode as a function of temperature and in the presence of
Allo or isoAllo. This absorption mode is in fact considered representative
of the acyl chain order in the bilayer, moving to higher wavenumbers if
the disorder increases [77]. The results and the detailed methods are reported in Supporting information. Briey, the insertion of Allo induces a
marked variation in the peak maximum in the direction of an increased
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
chemical components in the lipid bilayer which is able to shift the critical point will also affect the lipid rafts dynamics at constant temperature. Lipid rafts have been shown to play important roles in the
function of ligand-gated ion channels. Ligand-gated ion channels associated with lipid rafts will be affected in their behavior as a consequence
of alteration of the lipid raft lateral scale dimensions and characteristic
life times. In this context, in some reports it has been suggested that
the GABAA receptor is localized in lipid rafts [78] and, according to the
specic drug, the interactions with the receptors occur inside or outside
the lipid raft [79]. If lipid rafts spatial and time characteristics are
altered, the activity of the GABAA receptor will be affected and also its
sensitivity to drugs. The activity of the GABAA receptor has been demonstrated to depend on the cholesterol concentration, whose amount in a
bilayer signicantly alters its thermodynamic state. Furthermore, the effect of allosteric modulators of the GABAA receptors, such as Allo, changes depending on the cholesterol concentration in the membrane. For
example, we found that the effect of Allo and isoAllo on the phase distribution in the bilayer strongly depends on the phase situation of the
bilayer, hence on the cholesterol amount, being much more evident
when there is a strong unbalance between the two phases fractional
occupancy (see Supporting information).
Moreover, we showed that an isoform of Allo, which is reported to
inhibit functionally Allo effects, acts on the lipid bilayer in an opposite
way with respect to Allo. In fact, isoAllo induces a condensation of the
lipid bilayer and a relative increase of the Lo fraction. The effect of isoAllo
when compared to that elicited by the same concentrations of Allo is
smaller, probably due to a different partitioning in the membrane. It is
important to note that in Micropipette Aspiration experiments [57]
the relative area variation for the bilayer is also smaller in the case of
isoAllo than of Allo.
4. Conclusions
Lipid membrane properties are strongly dependent on their composition. In general, the different behaviors of Allo and isoAllo highlighted
in our work suggest that the effects of a drug on a lipid bilayer could
represent a mechanism able to modulate the activity of membrane proteins even when these proteins are activated by specic interactions
with the drug. In fact, considering the two hypothesis on the
Fig. 7. Effect of isoAllo on DOPC/bSM/chol 1:1:1 supported bilayer. ab) overall view of bilayer patches before a) and with 300 nM isoAllo in the imaging chamber b). cf) sequence of
images of lipid bilayer patches in the phase coexistence region as the concentration of isoAllo is increased: c) no isoAllo; d) 50 nM; e) 100 nM; f) 300 nM.
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
[25] G. Puja, M.R. Santi, S. Vicini, D.B. Pritchett, R.H. Purdy, S.M. Paul, P.H. Seeburg, E.
Costa, Neurosteroids act on recombinant human GABAA receptors, Neuron 4
(1990) 759765.
[26] J.M. Mienville, S. Vicini, Pregnenolone sulfate antagonizes GABAA receptor-mediated
currents via a reduction of channel opening frequency, Brain Res. 489 (1989)
190194.
[27] G. Akk, D.F. Covey, A.S. Evers, J.H. Steinbach, C.F. Zorumski, S. Mennerick, The inuence of the membrane on neurosteroid actions at GABAA receptors, Psychoneuroendocrinology 34S (2009) S59S66.
[28] S. Mennerick, Y. He, X. Jiang, B.D. Manion, M. Wang, A. Shute, A. Benz, A.S. Evers, D.F.
Covey, C.F. Zorumski, Selective antagonism of 5-reduced neurosteroid effect at
GABAA receptors, Mol. Pharm. 65 (2004) 11911197.
[29] A.M. Hosie, M.E. Wilkins, T.G. Smart, Neurosteroid binding sites on GABA(A) receptors,
Pharmacol. Ther. 116 (2004) 719.
[30] A.M. Hosie, L. Clarke, H. da Silva, T.G. Smart, Conserved site for neurosteroid
modulation of GABA A receptors, Neuropharmacology 56 (2009) 149154.
[31] M. Chisari, L.N. Eisenman, K. Krishnan, A.K. Bandyopadhyaya, C. Wang, A. Taylor, A.
Benz, D.F. Covey, C.F. Zorumski, S. Mennerick, The inuence of neuroactive steroid
lipophilicity on GABAA receptor modulation: evidence for a low-afnity interaction,
J. Neurophysiol. 102 (2009) 12541264.
[32] M.D. Majewska, Neurosteroids: endogenous bimodal modulators of the GABAA receptor. Mechanism of action and physiological signicance, Prog. Neurobiol. 38
(1992) 379395.
[33] H.J. Shu, L.N. Eisenman, D. Jinadasa, D.F. Covey, C.F. Zorumski, S. Mennerick, Slow
actions of neuroactive steroids at GABAA receptors, J. Neurosci. 24 (2004)
66676675.
[34] S.L. Veatch, S.L. Keller, A closer look at the canonical raft mixture in model membrane studies, Biophys. J. 84 (2003) 725726.
[35] H. Callachan, G.A. Cottrell, N.Y. Hather, J.J. Lambert, J.M. Nooney, J.A. Peters, Modulation of the GABAA receptor by progesterone metabolites, Proc. R. Soc. Lond. B Biol.
Sci. 231 (1987) 359369.
[36] M. Wang, Y. He, L.N. Eisenman, C. Fields, C.-M. Zheng, J. Mathews, A. Benz, T.
Fu, E. Zorumski, J.H. Steinbach, D.F. Covey, C.F. Zorumski, S. Mennerick, 3hydroxypregnane steroids are pregnenolone sulfate-like GABAA receptor antagonists, J. Neurosci. 22 (2002) 33663375.
[37] L.K. Tamm, H.M. McConnell, Supported phospholipid bilayers, Biophys. J. 47 (1985)
105113.
[38] H.M. McConnell, T.H. Watts, R.M. Weis, A.A. Brian, Supported planar membranes in
studies of cellcell recognition, Biochim. Biophys. Acta 864 (1986) 95.
[39] H.M. Seeger, G. Marino, A. Alessandrini, P. Facci, Effect of physical parameters on the
main phase transition of supported lipid bilayers, Biophys. J. 97 (2009) 10671076.
[40] S.M. Cook, T.E. Schffer, K.M. Lang, Practical implementation of dynamic methods for
measuring atomic force microscope cantilever spring constants, Nanotechnology 17
(2006) 21352145.
[41] I. Horcas, R. Fernndez, J.M. Gmez-Rodrguez, J. Colchero, J. Gmez-Herrero, A.M.
Baro, WSxM: a software for scanning probe microscopy and a tool for nanotechnology, Rev. Sci. Instrum. 78 (2007) 013705.
[42] S.L. Veatch, S.L. Keller, Seeing spots: complex phase behavior in simple membranes,
Biochim. Biophys. Acta 1746 (2005) 172185.
[43] S.L. Veatch, S.L. Keller, Miscibility phase diagrams of giant vesicles containing
sphingomyelin, Phys. Rev. Lett. 94 (2005) 148101.
[44] O.G. Mouritsen, The liquid order come of age, Biochim. Biophys. Acta 1798 (2010)
12861288.
[45] L.A. Bagatolli, Membrane domains and their relevance to the organization of the biological membranes, Compr. Biophys. 5 (2012) 1636.
[46] S.D. Connell, G. Heath, P.D. Olmsted, A. Kisil, Critical point uctuations in supported
lipid membranes, Faraday Discuss. 161 (2013) 91111.
[47] S.L. Veatch, I.V. Polozov, K. Gawrisch, S.L. Keller, Liquid domains in vesicles investigated by NMR and uorescence microscopy, Biophys. J. 86 (2004) 29102922.
[48] R.S. Petruzielo, F.A. Heberle, P. Drazba, J. Katsaras, G.W. Feigenson, Phase behavior
and domain size in sphingomyelin-containing lipid bilayers, Biochim. Biophys.
Acta 1828 (2013) 130213013.
[49] T. Hamada, Y. Kishimoto, T. Nagasaki, M. Takagi, Lateral phase separation in tense
membranes, Soft Matter 7 (2011) 9061.
[50] M. Fuhrmans, M. Muller, Mechanisms of vesicle spreading on surfaces: coarsegrained simulations, Langmuir 29 (2013) 43354349.
[51] M.R. Nussio, G. Oncins, I. Ridelis, E. Szili, J.G. Shapter, F. Sanz, N.H. Voelcker, Nanomechanical characterization of phospholipid bilayer islands on at and porous
substrates: a force spectroscopy study, J. Phys. Chem. B 113 (2009) 1033910347.
[52] T. Portet, S.E. Gordon, S.L. Keller, Increasing membrane tension decreases miscibility
temperatures; an experimental demonstration via micropipette aspiration, Biophys.
J. 103 (2012) L35L37.
[53] D. Chen, M.M. Santore, Large effect of membrane tension on the uidsolid phase
transitions of two-component phosphatidylcholine vesicles, Proc. Natl. Acad. Sci.
U. S. A. 111 (2014) 179184.
[54] J.E. Shaw, R.F. Epand, R.M. Epand, Z. Li, R. Bittman, C.M. Yip, Correlated uorescenceAtomic Force Microscopy of membrane domains: structure of uorescence probes
determines lipid localization, Biophys. J. 90 (2006) 21702178.
[55] S. Chiantia, J. Ries, N. Kahya, P. Schwille, Combined AFM and two-focus SFCS study of
raft-exhibiting model membranes, Chemphyschem 7 (2006) 24092418.
[56] J. Zhao, J. Wu, H. Shao, F. Kong, N. Jain, G. Hunt, G. Feigenson, Phase studies of model
biomembranes: macroscopic coexistence of L + L, with light induced coexistence
of L + Lo phases, Biochim. Biophys. Acta 1768 (2007) 27772786.
[57] D. Balleza, M. Sacchi, G. Vena, D. Galloni, G. Puia, P. Facci, A. Alessandrini, Effects of
neurosteroids on a model membrane including cholesterol: A Micropipette Aspiration study, 2015. (submitted for publication).
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002
10
[58] T. Bhatia, P. Husen, J.H. Ipsen, L.A. Bagatolli, A.C. Simonsen, Fluid domain patterns in
free-standing membranes captured on a solid support, Biochim. Biophys. Acta 1838
(2014) 25032510.
[59] M. Gedig, S. Faiss, A. Janshoff, Melting and interdigitation of microstructured solid
supported membranes quantied by imaging ellipsometry, Biointerphases 3
(2008) FA51.
[60] S.L. Veatch, S.S. Leung, R.E. Hancock, J.L. Thewalt, Fluorescent probes alter miscibility
phase boundaries in ternary vesicles, J. Phys. Chem. B 111 (2007) 502504.
[61] T. Sooksawate, M.A. Simmonds, Inuence of membrane cholesterol on modulation
of the GABAA receptor by neuroactive steroids and other potentiators, Br. J.
Pharmacol. 134 (2001) 13031311.
[62] T. Sooksawate, M.A. Simmonds, Increased membrane cholesterol reduces the
potentiation of GABAA currents by neurosteroids in dissociated hippocampal
neurons, Neuropharmacology 37 (1998) 11031110.
[63] T. Sooksawate, M.A. Simmonds, Effects of membrane cholesterol on the sensitivity
of the GABAA receptor to GABA in acutely dissociated rat hippocampal neurons,
Neuropharmacology 40 (2001) 178184.
[64] L.S. Hirst, P. Uppamoochikkal, L. Chai, Phase separation and critical phenomena in
biomimetic ternary lipid mixtures, Liq. Cryst. 38 (2011) 17351747.
[65] A.R. Honerkamp-Smith, P. Cicuta, M.D. Collins, S.L. Veatch, M. den Nijs, M. Schick, S.L.
Keller, Line tensions, correlation lengths, and critical exponents in lipid membranes
near critical points, Biophys. J. 95 (2008) 236246.
[66] J.H. Davis, M.L. Schmidt, Critical behavior in DOPC/DPPC/cholesterol mixtures: Static
2H NMR line shapes near the critical point, Biophys. J. 106 (2014) 19701978.
[67] A. Alessandrini, P. Facci, Atomic force microscope: a versatile tool in biophysics,
Meas. Sci. Technol. 16 (2005) R65R92.
[68] A. Alessandrini, P. Facci, Phase transitions in supported lipid bilayers studied by
AFM, Soft Matter 10 (2014) 71457164.
[69] L.L. Wittmer, Y. Hu, M. Kalkbrenner, A.S. Evers, C.F. Zorumski, D.F. Covey,
Enantioselectivity of steroid-induced gamma-aminobutyric acidA receptor modulation and anesthesia, Mol. Pharmacol. 50 (1996) 15811586.
[70] K.R. Nilsson, C.F. Zorumski, D.F. Covey, Neurosteroid analogues. 6. The synthesis and
GABAA receptor pharmacology of enantiomers of dehydroepiandrosterone sulfate,
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
[79]
Please cite this article as: M. Sacchi, et al., Effect of neurosteroids on a model lipid bilayer including cholesterol: An Atomic Force Microscopy
study, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.01.002